Making AI Adoption ROI Transparent: Empirical Dispersion, Payback Lags, and the Productivity Paradox
This paper analyses 200,000 firm-year ROI observations across 8,000 large enterprises to examine why AI productivity gains frequently fail to convert into positive financial returns. While 66.5% of observations achieved productivity gains of 30% or more, only 24.8% recorded positive same-year ROI. Using breakeven theory and the law of diminishing returns, the study identifies two breakeven points and shows that top-line revenue conversion, rather than capital spend or internal efficiency, is the primary driver of ROI profitability. Financial payback concentrates in Technology, Financial Services and Retail. The findings challenge the assumption that higher AI adoption and training automatically produce positive ROI demonstrating that productivity gains do not always convert to ROI.
Authors
- Karen Laing
- Rex Orchid Labs (ORCID: https://orcid.org/0009-0005-8451-3261)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-21
- DOI
- https://doi.org/10.5281/zenodo.24408359
- Primary Topic
- Impact of AI and Big Data on Business and Society
- Type
- preprint